One observer was assigned to the task of determining specificity. This was accomplished by interspersing normal and fibrocystic cases between cancer cases. His accuracy was 90%, although most cases missed were errors of interpretation rather than failure of informational content in the film.
Cancer size varied from 3-mm area of microcalcifications to 6-cm neoplasms. Over 90% of cases were new; 10% had previous films for comparison.